were available from fish released as fry. 



The three mean scale graphs for the marked 

 and unmarked female fall chinook salmon that 

 returned to the Spring Creek Hatchery in 1960 

 as 4-year-olds are shown in figure 6. These 

 graphs show the growth period from the begin- 





IS 

 ■s 



MARKED. SIZE !]-)( INCHES 



N=I7 

 UNMARKED. SIZE !S- 36, INCHES 



N=!S 

 UNMARKED. SIZE ))-H INCHES 



N=I7 



"T" 



"T" 



"T" 



1 2 3 4 S 6 7 



DISTANCE EROM FOCUS ALONG ANTERO- LATERAL 



AXIS. IN 20-MM- UNITS (X 140) 



Figure 6. — Mean scale graphs of marked and unmarked 

 female fall chinooks of three selected size groups that 

 returned to Spring Creek Hatchery in 1960 as 4-year- 

 old fish. 



ning of scale growth toward the end of the first 

 summer in the ocean. They represent, there- 

 fore, only part of the antero-lateral radius. The 

 initial part of the graphs (units 1-4), which 

 represents fresh-water and intermediate 

 growth, is very similar among the three groups. 

 The remaining part of the graphs (units 5-8), 

 which represents the major part of first marine 

 growth, becomes divergent in that the mean 

 circulus spacing of the marked group is con- 

 sistently smaller than that of either of the 

 two unmarked groups (33-34 inches, 35-36 

 inches), with greater diff'erence shown between 

 the marked and the larger unmarked fish. 



The fact that the initial part of the scale 

 graphs is similar among the three groups of 

 chinook salmon is easily understood, because 

 marking is applied during the fingerling stage 

 after the fish have grown the initial part of 

 the scale. The diff'erence in the remaining part 

 of the scale graphs between the marked and 

 unmarked groups, especially between those of 

 the same size, strongly suggests that marking 



has slowed down the fish's growth rate, at least 

 during the first summer in the ocean. 



To verify the results revealed by the scale 

 graphs, the total distance from the 16th to 25th 

 circulus, which represents the major part of 

 the first year's marine growth, was measured 

 and a f-test applied (table 2). We first tested 

 sample variances, and in both pairs equality 



Tapi.e 2. — Frrrpirnr;/ anil xtnli^ticx of mcasuremenix of scale 

 ijrnu'th (iurimi first i/ear's iiKiririf life in ni'irkcil anil unmarked 

 fall chinook salmon that returned in 1060 to Spring Creek 

 Hatchery as /^-year-olds 



Distance 

 of 10 circiili 

 (lBtll-2Sth) 



31. 

 32. 

 33. 

 34. 

 35. 

 36. 

 37. 

 38.. 

 39.. 

 40.. 

 41.. 

 42.. 

 43.. 

 44.. 

 45.. 

 46.. 

 47.. 

 48.. 

 49.. 

 50.. 

 51.. 

 52.. 

 53.- 

 54.. 

 55.. 



Mm. z no 



s- 



Q."! percent con- 

 fidence limits 

 for ratio of 

 population 

 variances. 



(■statistic 



(If 



V.iliicnf [(I at 

 n.(),'> siRiiifi- 

 cance level. 



9.T percent confi- 

 dence limits 

 for (lifTerence 

 lir'tueen 

 population 

 means. 



Marked 



fish 33-34 



inches long 



Nu mber 



I'mnarked 



fish 33-34 



inches long 



Number 



Marked 



fish 33-34 



inches long 



Number 



Unmarked 



fish 3.i-3« 



inches long 



Number 



17 



40.88 



30.23 



4.'i.47 

 23.89 



0.29<-'A;<2.I9 

 SB- 



-3.58* 

 32 



2.04 



-9.95<;i,-;i!<-2.73 



17 



40.88 



30.23 



4.12* 



26 



47.27 



22.12 



41 



2.02 

 3.25<^i-jii<9..'i3 



can be accepted ; therefore, a simple t-test was 

 used. The width of ten marine circuli was sig- 

 nificantly greater in the unmarked groups than 

 in the marked group, compared either between 

 two modal lengths of fish or between fish of the 

 same length. These tests thus confirm the re- 

 sults obtained by the scale graph method. 



172 



U.S. FISH AND WILDLIFE SERVICE 



